A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 23:45 UTC
Revert "[AMD][CI/Build] Fix the AMD issue caused by inappropriate of symbol exposure (#21647)" (#21850)
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
Gregory Shtrasberg committed
Jul 29, 2025 at 17:56 UTC
176bbce1db0ba81230c396cb46cf4035a16d2c66
1 file changed
+2
-2
CMakeLists.txt
+2
-2
@@ -243,6 +243,7 @@ set(VLLM_EXT_SRC
243
"csrc/sampler.cu"
244
"csrc/cuda_view.cu"
245
"csrc/quantization/gptq/q_gemm.cu"
246
+ "csrc/quantization/compressed_tensors/int8_quant_kernels.cu"
247
"csrc/quantization/fp8/common.cu"
248
"csrc/quantization/fused_kernels/fused_layernorm_dynamic_per_token_quant.cu"
249
"csrc/quantization/gguf/gguf_kernel.cu"
@@ -296,8 +297,7 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
297
"csrc/sparse/cutlass/sparse_scaled_mm_entry.cu"
298
"csrc/cutlass_extensions/common.cpp"
299
"csrc/attention/mla/cutlass_mla_entry.cu"
299
- "csrc/quantization/fp8/per_token_group_quant.cu"
300
- "csrc/quantization/compressed_tensors/int8_quant_kernels.cu")
300
+ "csrc/quantization/fp8/per_token_group_quant.cu")
301
302
set_gencode_flags_for_srcs(
303
SRCS "${VLLM_EXT_SRC}"